dissenting.
When a judge of a court of appeals has written a majority opinion for the court1 and finds after an en banc rehearing that he not only has no votary among the other active judges of the court but, indeed, he has no reluctant followers, he would be foolish if he did not assiduously reexamine the pinions of his previously held opinion. I have done so and have concluded that to abandon what I sincerely believed to be a proper and required result would be also an abandonment of my duty as a judge. I do not launch on the preparation of a dissent with any feeling of quixotic disregard for reality for I regard the result reached by the majority opinion as eviscerating a fundamental principle of our constitution, the taking of property without compensation. The result reached in this case may be pleasing to the zealous believers in a fresh start for bankrupts but I cannot believe it has been realized without sacrificing a basic concept. I therefore respectfully dissent.
As the majority properly notes, the “first question” in this case is whether Congress intended section 522(f)(2) to apply to security interests that attached prior to November 6, 1978, the enactment date of the Bankruptcy Reform Act of 1978. I am persuaded that retrospective application of section 522(f)(2) would give rise to grave constitutional questions under the takings *461clause of the Fifth Amendment and that the statute should therefore be construed to apply prospectively only, in light of the strong policy mandate, reemphasized by the Supreme Court in NLRB v. Catholic Bishop, 440 U.S. 490, 99 S.Ct. 1313, 59 L.Ed.2d 533 (1979), that courts construe statutes in a manner that, if possible, avoids constitutional questions. See, e.g., Crowell v. Benson, 285 U.S. 22, 62, 52 S.Ct. 285, 296-97, 76 L.Ed. 598 (1932); Murray v. The Schooner Charming Betsy, 6 U.S. (2 Cranch) 64, 118, 2 L.Ed. 208 (1804); 2A C. Sands, Sutherland Statutory Construction § 45.11 (4th Ed. 1973). Indeed, the Court has repeatedly emphasized “that the words of a statute may be strained ‘in the candid service of avoiding a serious constitutional doubt.’ United States v. Rumely, 345 U.S. 41, 47 [73 S.Ct. 543, 546, 97 L.Ed. 770] (1953).” United States v. Seeger, 380 U.S. 163, 188, 85 S.Ct. 850, 865, 13 L.Ed.2d 733 (Douglas, J., concurring); Blodgett v. Holden, 275 U.S. 142, 148, 48 S.Ct. 105, 107, 72 L.Ed. 206 (1927) (Holmes, J., concurring).
In NLRB v. Catholic Bishop, 440 U.S. 490, 99 S.Ct. 1313, 59 L.Ed.2d 533 (1979), the Supreme Court emphasized the proper mode of analysis of a Congressional enactment which is claimed to be unconstitutional. The Catholic Bishop Court had before it the question of whether lay teachers in church-operated schools were within the jurisdiction of the NLRB. The Court found that neither the language nor the legislative history of the National Labor Relations Act disclosed “an affirmative intention ... clearly expressed,” that the NLRB have such jurisdiction. Absent such clear expression of Congressional intent to bring teachers within the NLRB’s jurisdiction, the Court declined to construe the Act in a manner that would require the resolution of “difficult and sensitive questions arising out of the guarantees of the First Amendment Religion Clauses.” 440 U.S. at 507, 99 S.Ct. at 1322. We turn first, therefore, to examination of whether retrospective application of § 522(f)(2) would give rise to such serious constitutional questions under the Fifth Amendment.
I.
The Constitution confers broad powers upon Congress to establish “uniform Laws on the subject of Bankruptcies.” U.S. Const. Art. I, § 8, cl. 4. Congress has acted under this broad grant to enact various statutes which have applied to rights which predated their enactment. See, e.g., Hanover National Bank v. Moyses, 186 U.S. 181, 22 S.Ct. 857, 46 L.Ed. 1113 (1902) (Bankruptcy Act of 1898 discharged plaintiff’s 1892 judgment against debtor); In re Prima Co., 88 F.2d 785 (7th Cir. 1937) (Congress intended Bankruptcy Code to confer on bankruptcy courts the power to issue trustee certificates with priority over existing mortgages).
The Bankruptcy Reform Act of 1978, a comprehensive revision of the entire bankruptcy system, was enacted in the exercise of this broad bankruptcy power. It applies to many rights and transactions which took place before its enactment, saving only those cases commenced under the old bankruptcy law. See, e.g., 11 U.S.C. § 365(b)(2) (invalidates bankruptcy clauses in executory contracts and unexpired leases); 522(e) (makes contractual waivers of exemptions unenforceable by unsecured creditors); 524(c) (contractual reaffirmations of debts unenforceable in some circumstances).
Courts have acknowledged the scope of the bankruptcy power as extending to all legislation regarding the discharge of contractual debts and distribution of the debt- or’s assets. See, e.g., Kuehner v. Irving Trust Co., 299 U.S. 445, 451, 57 S.Ct. 298, 301, 81 L.Ed. 340 (1937); Hanover National Bank, 186 U.S. at 186, 22 S.Ct. at 860. It is also settled, however, that the bankruptcy power is not without limitation, but rather is subject to the constitutional limits of the Fifth Amendment. Louisville Joint Stock Land Bank v. Radford, 295 U.S. 555, 589, 55 S.Ct. 854, 863, 79 L.Ed. 1593 (1935); Rodrock v. Security Industrial Bank, 642 F.2d 1193, 1197 (10th Cir. 1981), prob. juris, noted, 454 U.S. 1122, 102 S.Ct. 969, 71 L.Ed.2d 108; In re Penn Central Transportation Co., 494 F.2d 270, 278 (3d Cir. 1974), cert. denied, *462419 U.S. 883, 95 S.Ct. 147, 42 L.Ed.2d 122. The focus of our inquiry at this juncture must be whether application of § 522(f)(2) to preenactment liens presents a significant risk of overstepping those constitutional limitations.
In Louisville Joint Stock Land Bank v. Radford, 295 U.S. 555, 55 S.Ct. 854, 79 L.Ed. 1593 (1935), the Supreme Court found that the Frazier-Lemke Act of 1934, which by its terms applied retroactively, conflicted with the Fifth Amendment, and was therefore void. The Act was designed to save family-owned farms from foreclosure by permitting a debtor to obtain a five-year stay of state foreclosure proceedings, to remain on the farm during the stay if he paid a reasonable rent, and to satisfy the mortgagee’s claim at any time within the five-year period by paying the mortgagee the appraised value of the property.
The Radford Court noted the broad power of Congress to discharge a debtor’s personal obligations, but concluded:
the effect of the Act here complained of is not the discharge of . . . personal obligation. It is the taking of substantial rights in specific property acquired by the Bank prior to the Act.
The province of the Court is limited to deciding whether the . . . Act ... as applied has taken from the Bank without compensation, and given to Radford, rights in specific property which are of substantial value.... As we conclude that the Act as applied has done so, we must hold it void. For the Fifth Amendment commands that, however great the Nation’s need, private property shall not thus be taken even for a wholly public use without just compensation.
295 U.S. at 589-90, 601-02, 55 S.Ct. at 863-64, 868-69 (citations omitted).2
The bankruptcy courts which have considered the constitutionality of section 522(f)(2) have split on the constitutionality of retrospective application. A substantial number have relied on the Radford rationale to find that section 522(f)(2) constitutes a taking. See, e.g., Glynn v. Household Finance Corp., 13 B.R. 647 (D.S.C.1981); Carroll v. Household Finance Corp., 11 B.R. 45 (Bkrtcy.E.D.N.Y.1981); In re Cunningham, 17 B.R. 463 (Bkrtcy.W.D.Ky.1981); In re Coleman v. American Finance Corp., 10 B.R. 772 (Bkrtcy.D.Md.1981). A similar number have upheld the validity of the section. See, e.g., Hinson v. Lexington State Bank, 20 B.R. 753 (Bkrtcy.D.S.C. 1982); Lucas v. Beneficial Finance Co., 18 B.R. 179 (Bkrtcy.S.D.Ohio 1982); Safeway Finance Co. v. Ward, 14 B.R. 549 (S.D.Ga. 1981); Curry v. Associates Financial Services, 11 B.R. 716 (N.D.Ohio 1981); Clark v. Savings & Trust Co., 11 B.R. 828 (Bkrtcy.W.D.Pa.1981). The courts of appeals which have considered the question have split on the constitutionality of retroactive application. Rodrock v. Security Industrial Bank, 642 F.2d 1193 (10th Cir. 1981), prob. juris, noted, 454 U.S. 1122, 102 S.Ct. 969, 71 L.Ed.2d 108 (unconstitutional); In re Ashe, 669 F.2d 105 (3d Cir. 1982) (§ 522(f)(1) constitutional); see Webber v. Credithrift, 674 F.2d 796 (9th Cir. 1982) (Act constitutional when applied to postenactment, preeffective liens; implies unconstitutional if applied to preenactment liens).
Of the courts upholding the constitutionality of the section, most have analyzed the section not in terms of a taking, but rather have relied on a highly expansive conception of the bankruptcy power, and a due *463process, rather than a takings analysis.3 See, e.g., Ashe, 669 F.2d at 110-11, see Note, Constitutionality of Retroactive Lien Avoidance under Bankruptcy Code Section 522(f), 94 Harv.L.Rev. 1616, 1621 & nn. 36-42 (1981). I am persuaded that the better reasoned analysis focuses on the taking clause and the Radford case.
In Rodrock v. Security Industrial Bank, 642 F.2d 1193 (10th Cir. 1981), prob. juris, noted, 454 U.S. 1122, 102 S.Ct. 969, 71 L.Ed.2d 108, the Court of Appeals for the Tenth Circuit analyzed section 522(f)(2) under the Radford case. The court reasoned:
In the instant cases, the creditors acquired rights in specific property prior to the enactment of the Reform Act, and, under Radford, these vested rights cannot be taken from the creditor for the benefit of the debtor. It should be noted that, in the instant eases, there would be a complete taking of the secured creditors’ property interests.
642 F.2d at 1197. The Rodrock court was also faced with the argument raised by the debtors here, that Radford has been discredited by a series of subsequent Supreme Court decisions. As the Giffords point out, Radford did not expressly rely on the taking clause of the Fifth Amendment, although the language of the opinion of the Court would seem so to indicate. Subsequent Supreme Court cases have seemed to treat Radford as a substantive due process case, and are silent on the taking issue. Wright v. Union Central Life Insurance Co., 311 U.S. 273, 61 S.Ct. 196, 85 L.Ed. 184 (1940); Wright v. Vinton Branch, 300 U.S. 440, 57 S.Ct. 556, 81 L.Ed. 736 (1937); see Helvering v. Griffiths, 318 U.S. 371, 400-01 n.52, 63 S.Ct. 636, 652 n.36, 87 L.Ed. 843 (1943). While these later cases do indeed circumscribe the application of Radford, and erode the substantive due process test it arguably applied, it is also clear that, “The Supreme Court .. . has never repudiated the principle that the value of a security interest in specific property is a fifth amendment property right that cannot be taken unless just compensation is given.” Note, 94 Harv.L.Rev., supra, at 1624. Note, The Continuing Vitality of Louisville Joint Stock Land Bank v. Radford, 15 Ind.L.Rev. 593 (1982).
The Supreme Court has continued to cite Radford for the proposition that the Government may not take a property interest without just compensation. See, e.g., Armstrong v. United States, 364 U.S. 40, 44, 80 S.Ct. 1563, 1566, 4 L.Ed.2d 1554 (1960). It is also worthy of note, although perhaps not conclusive, that Congress apparently considered Radford still vital at the time of the passage of the Bankruptcy Reform Act. S.Rep.No. 95-989, 95th Cong., 2d Sess. 49, 76, reprinted in [1978] U.S.Code Cong. & Ad.News 5787, 5835, 5862 (citing Radford with approval); H.R.Rep.No. 95-595, 95th Cong., 2d Sess. 339, 361, reprinted in [1978] U.S.Code Cong. & Ad.News 5963, 6295, 6317. We therefore agree with the Rodrock court that subsequent Supreme Court decisions, “may well refine .. ., but ... do not destroy the fundamental teaching of Rad-ford that Congress may not under the bankruptcy power completely take for the benefit of a debtor rights in specific property previously acquired by a creditor.” 642 F.2d at 1198.
As Rodrock makes clear, the continuing vitality of Radford raises substantial and significant Fifth Amendment questions concerning retrospective application of section 522(f)(2). Because I see no escape from the question whether a taking has occurred if section 522(f)(2) is applied to liens which attached under state law to specific property prior to the enactment date of the Act, I believe it is appropriate to turn to an examination of the statute- to decide whether it was the affirmatively expressed intent of Congress that the statute be read to apply retrospectively and thus require resolution *464of the constitutional question, or whether it can be construed to apply prospectively only.
II.
As the majority opinion correctly points out, section 522(f)(2) does not state whether it is to apply retrospectively. As is the case with the rest of the new Act, the only express temporal limits on application of the section are found in the transition provisions of section 401 and 402(a), which except from the coverage of the new Act only those cases commenced prior to October 1, 1979, the effective date of the new Act. I do not, however, view this as dispositive of the question whether Congress intended section 522(f)(2) to apply to preenactment liens.4 As the Supreme Court pointed out long ago:
frequently words of general meaning are used in a statute, words broad enough to include an act in question, and yet a consideration of the whole legislation, or of the circumstances surrounding its enactment, or of the absurd results which follow from giving such broad meaning to the words, makes it unreasonable to believe that the legislator intended to include the particular act.
Church of the Holy Trinity v. United States, 143 U.S. 457, 459, 12 S.Ct. 511, 512, 36 L.Ed. 226 (1892). The Holy Trinity Court held that where reliance on the “plain meaning” of the statute led to an absurd result, resort to legislative history and other extrinsic aids to construction is appropriate, and their interpretative value may be controlling. In this case, where application of the general rule of sections 401 and 402(a) to section 522(f)(2) would lead to a result I believe would not be merely absurd, but in violation of the Constitution, resort to such aids to ascertain whether such a result was indeed the intent of the legislature is not only advisable, it is compelled. See, e.g., United States v. Clark, 445 U.S. 23, 27-31, 100 S.Ct. 895, 899-901, 63 L.Ed.2d 171 (1980). I turn therefore, to examine the legislative history of the 1978 Act to determine whether its broad sweep was intended to include the avoidance of preenactment liens.
First, changes in the preliminary drafts of the transition provisions suggest that Congress did not intend retrospective application of section 522(f)(2). As the majority opinion points out, preliminary drafts of the sections expressly required retrospective application. William Plumb, Jr., a consultant to the Commission on Bankruptcy Law of the United States, pointed out in hearings before the House that such a provision could be viewed as an improper impairment of property rights if given retrospective effect. Bankruptcy Act Revision: Hearings on H.R. 31 and H.R. 32 Before the House Subcomm. on Civil and Constitutional Rights, 94th Cong., 1st Sess. 2034, 2066-67 (1976). The provision requiring retroactivity was then stricken from the language of the transition provisions, and the present *465language inserted in its place. I find this persuasive evidence that Congress recognized the Fifth Amendment limitations on the Bankruptcy power, and sought to stay within them. Cf. Bradley v. School Board, 416 U.S. 696, 716 n.23, 94 S.Ct. 2006, 2018 n.23, 40 L.Ed.2d 476 (1974) (when Congress struck language limiting the award of attorneys’ fees to fees incurred after the statute’s effective date, the Court would not read such a limitation back into the statute). The majority’s observation that Mr. Plumb was only one of many witnesses testifying before Congress does not dilute either the correctness of his conclusions, or the evidentiary force of the deletion of the retroactivity provisions.
Second, the absence of any express indication in the legislative history of section 522(f)(2) that Congress intended the section to apply to vested property rights in existence at the time of enactment is more evidence of Congressional sensitivity to the limits of the Bankruptcy power. Further demonstration of this recognition is reflected by repeated references to the Rad-ford decision in both the House and Senate reports accompanying the Act, as noted previously herein.5 I therefore conclude that there was no Congressional “affirmative intention . . . clearly expressed,” Catholic Bishop, 440 U.S. at 507, 99 S.Ct. at 1322, that the Act apply to preenactment security interests.
Having determined that retroactive application of the statute would give rise to serious constitutional questions, and that there is no specific Congressional mandate that the section be so applied, I would construe the statute to apply prospectively only, i.e., only to those liens which attached after the enactment date6 of the statute, *466and conclude my analysis there.7 As author of the opinion of this court in NLRB v. Catholic Bishop, 559 F.2d 1112 (7th Cir. 1977), aff’d on different grounds, 440 U.S. 490, 99 S.Ct. 1919, 59 L.Ed.2d 533 (1979), I am particularly aware of the Supreme Court’s preference for reaching a decision on other than constitutional grounds. Because the entire thrust of the majority opinion is that the statute as applied to preenactment liens is constitutional, however, I too will proceed to address the constitutional questions such application presents.
III.
Although the majority opinion correctly sets forth the analytical factors the Supreme Court has established for a non-physical invasion8 takings case, I believe its attempt to apply those factors in a manner that distinguishes the Radford case9 falls short of the mark. To rephrase the majority, the three key factors in a non-physical invasion takings case are: (1) the quality of the property interest asserted, or “the investment-backed expectation”; (2) the economic impact of the Governmental action; and (3) the character of the Governmental action. The majority opinion erroneously concludes that the creditor’s expectation is illusory, the economic impact of the Governmental action is insubstantial, and that the Governmental action does not rise to the level of a taking because it is rational, not a physical invasion of the property, and does not inure to the Government’s own benefit, but is rather “ordinary economic regulation,” immune from a takings challenge. Because I am convinced that the majority incorrectly analyzes each of these three elements of the takings test, I respectfully dissent, and would conclude that retrospective application of section 522(f)(2) is unconstitutional and the statute is invalid to the extent it is so applied.
A. The property interest affected
A lien or security interest in personal property gives a secured party substantial rights under Wisconsin law, including:
(1) the right to take possession and dispose of the collateral upon default; Wis. Stat.Ann. § 409.503 (West 1964);
(2) the rights to have the collateral sold, publicly or privately, and to buy the collateral at such sale; id. at § 409.504;
(3) the right to retain the collateral in satisfaction of the debt; id. at § 409.505.
It is well settled under state and federal law, see 11 U.S.C. § 101(28) (Supp. III 1978), that these rights, essentially indistinguishable from those adumbrated in Rad-ford, see n.2, supra, constitute a “property interest” in the collateral to which they attach.
*467That such rights constitute a property interest rather than a contractual right to repayment is not merely an “academic difference,” as the majority suggests. It is a fundamental difference in kind, and of constitutional significance. Kuehner v. Irving Trust Co., 299 U.S. 445, 451-52, 57 S.Ct.-298, 301, 81 L.Ed. 340 (1937) (“there is, as respects the exertion of the bankruptcy power, a significant difference between a property interest and a contract, since the Constitution does not forbid impairment of the obligation of the latter”). Kuehner pointed out that where a property right exists, the Radford holding remains controlling. The majority attempts to avoid that holding by distinguishing Thorp’s property interest in the Gifford’s property from that in Radford on two grounds: (1) the security interest in the instant case is not a purchase money security interest; and (2) the fair market value of the collateral is “insignificant,” and the lien’s value, therefore, inheres primarily in the threat of foreclosure. These differences do not factually distinguish this case from Radford. Even if one accepts that such a distinction may be made, it does not rise to a constitutionally significant level allowing the extinction of Thorp’s property right.
Turning first to the purchase money/non-purchase money distinction, I note that Radford itself did not involve a loan given to purchase a farm, but rather involved a second mortgage on the Radford’s equity in their farm. 295 U.S. at 573-74 & nn. 3 & 4, 55 S.Ct. at 855-56 nn. 3 & 4. The majority’s procrustean attempt to characterize the Radford loan as a purchase money loan on the grounds that it was used to purchase tools and seed which were used to benefit the burdened property is untenable. First, I find no indication in Radford as to the type of goods on which the loan was spent, and the majority’s “seed and tools” hypothesis is thus entirely a matter of speculation. Second, the “used to benefit the property” distinction is essentially a matter of semantics depending on a broad definition of benefited property:10 the characterization could as easily be applied to the instant case in which the proceeds of the loan benefited the entire Gifford household. Under the majority analysis, therefore, the attachment of liens to items of household goods, which after all are only a portion of the benefited property, should be acceptable to the same degree as in Radford. In neither ease, however, can this be said to convert the lien to a purchase money security interest.11 Thus this case simply cannot be distinguished from Radford on the basis of any purchase money/nonpurchase money dichotomy.12
The majority then attempts to distinguish this case from Radford on the basis that the market value of the collateral is insignificant, and only the “extortion value” of the security interest makes it of value to Thorp. I note initially that the fact that the market value of the collateral is substantially less *468than the debt it secures, again, does not distinguish the case factually from Radford. In Radford, the farm had a market value of $18,000 when credit of $9,000 was extended. At bankruptcy, however, the farm’s value had deteriorated to $4,445. 295 U.S. at 573, 577, 55 S.Ct. at 855, 857. Thus foreclosure would not have resulted in complete satisfaction of the debtor’s obligation. This did not justify the destruction of what security remained in Radford, and it cannot justify it here. See Armstrong v. United States, 364 U.S. 40, 80 S.Ct. 1563, 4 L.Ed.2d 1554 (1960) (unless secured claim is “valueless,” it cannot be eliminated).
The heart of the majority’s argument is the assertion that because the real value of the liens lies in the threat of foreclosure, there would be no justice in compensating Thorp for that element of value. Even if we should assume that the lenders who take a security interest in leasehold goods never intend to take possession of the tangible items (a fact which Thorp denies) but the only purpose of securing the lien is as leverage on the borrower to repay the money borrowed, it is departing from the real world to say that we are not dealing with valuable assets. The very purpose of any collateral whether it be in connection with a purchase-money transaction or be it a farm or a Mercedes sport car is to provide a compulsion for the borrower to repay. The retroactive application of section 522(f) is a taking of a valuable property and to argue otherwise is to ignore the basis on which a multibillion dollar lending business has operated. Thorp’s preference of repayment to foreclosure does not, therefore, distinguish it from other creditors in any fundamental sense. All creditors take security to insure repayment, not primarily to substitute collateral for repayment. The fact that section 522(f)(2) creditors are perhaps more interested in repayment than in the taking possession of collateral than other creditors might be, although I doubt this is true,13 distinguishes them only in degree, and cannot support the majority’s constitutional distinction. See Note, 94 Harv.L.Rev., supra, at 1632.
The majority relies heavily on Congressional findings for the conclusion that such liens are rarely foreclosed upon, but rather are used as tools of coercion. Characterizing legally sanctioned rights as oppressive, however, does not entitle Congress to abrogate the protections of the Fifth Amendment. Excessive judicial deference to such Congressional definitions of what is and is not a property right would eviscerate the takings clause of the Fifth Amendment.
I am convinced, for all the above reasons, that the creditor’s property interest here is indistinguishable from that involved in the Radford case. Thorp’s expectation 14 was *469that it could retain its lien until the Giffords paid their debt, and that if they defaulted it would have certain specific rights under state law. It backed that reasonable expectation with an investment of some $3,000. I would hold that this satisfies the first element of the takings analysis.15
B. The economic impact of the section
The Supreme Court has described this aspect of the takings analysis as based upon “the extent to which the regulation has interfered with distinct investment-backed expectations.” Penn Central Transportation Co. v. New York, 438 U.S. 104, 124, 98 S.Ct. 2646, 2659, 57 L.Ed.2d 631 (1978). As the Rodrock court pointed out, section 522(f)(2) results in “a complete taking of the secured creditors’ property interest.” 642 F.2d at 1197. It substitutes the right of an unsecured creditor to share in the distribution of the debtor’s remaining assets — a fundamentally different right of highly dubious worth in most consumer bankruptcies, see, e.g., Kuehner, 299 U.S. at 451-52, 57 S.Ct. at 301-02; Rodrock, 3 Bankr.Rptr. at 630 & n.2 (D.Colo.1980); Note, 94 Harv.L. Rev., supra, at 1631. This total obliteration of any property right in the collateral is the ultimate in impact. In my view, it satisfies the second element of the takings test. The majority, however, finds the impact de minimis on three grounds: (1) since the lien can only be avoided up to $200 per item, it has a minimal impact on creditors with truly valuable security; (2) a creditor’s expectation of repayment is at a minimum once bankruptcy occurs; and (3) substitution of the rights of an unsecured creditor mitigates whatever impact the section has.
I am unwilling to accept, even in this inflationary age, the notion that $200 is per se a de minimis amount. Furthermore, although the worth of any given item of household goods may be under $200, the aggregate amounts involved in such transactions are considerably larger. Finance companies such as Thorp made 79,720 consumer loans in Wisconsin in 1978, totalling $148,408,699. Consumer Credit Division, Office of the Commissioner of Banking of Wisconsin.16 Thus the dollars and cents impact of the majority’s interpretation is substantial in the extreme.
Nor am I moved by the majority’s contention that a secured creditor’s expectations are at a minimum in bankruptcy. This may well be true of the unsecured creditor, but the secured creditor has always had the worth of his collateral or the collateral itself to fall back upon in case either of default or bankruptcy. See Wis. Stat.Ann. § 409.501 (West 1964) (Official *470UCC Comment: “The rights of the secured party in the collateral after the debtor’s default are of the essence of a security transaction. These are the rights which distinguish the secured from the unsecured lender”); Long v. Bullard, 117 U.S. 617, 6 S.Ct. 917, 29 L.Ed. 1004 (1886) (secured creditor’s lien protected in bankruptcy). Thus only if we assume the majority’s conclusion that 522(f)(2) applies would the creditor’s expectation of retaining that protection vanish. I reject this essentially circular argument, as well as the contention that Thorp could have avoided this loss by careful monitoring of its accounts or renegotiation of the loan. There is no indication in the record that the Giffords defaulted prior to bankruptcy. Thorp had no rights under state law absent such default to control the collateral or the Giffords’ spending. While the renegotiation for other collateral suggestion sounds appealing, I question its plausibility. What bargaining power would Thorp have had to seek additional or substitute collateral in the absence of any extension of additional credit; and if it felt additional collateral was needed because of the risk of default such an extension would certainly seem to have been commercially foolhardy.
Nor does the somewhat frivolous suggestion of the substitution of the rights of an unsecured creditor mitigate the impact of section 522(f)(2) on Thorp’s property interest. As noted above, an unsecured claim differs fundamentally in kind and in value from a secured claim. An unsecured claim in a consumer bankruptcy is generally worthless.17 See supra. Substitution of these utterly illusory rights has the same effect as complete destruction of the property right, and cannot mitigate the section’s economic impact. See Pennsylvania Coal Co. v. Mahon, 260 U.S. 393, 414-15, 43 S.Ct. 158, 159-60, 67 L.Ed. 322 (1922).
C. The character of the Government action
Finally, I am persuaded that the character of the Government action fulfills the requirements of this final element of the takings analysis.18 The recent cases of Kaiser Aetna v. United States, 444 U.S. 164, 100 S.Ct. 383, 62 L.Ed.2d 332 (1979); Penn Central Transportation Co. v. New York, 438 U.S. 104, 98 S.Ct. 2646, 57 L.Ed.2d 631 (1978); and Andrus v. Allard, 444 U.S. 51, 100 S.Ct. 318, 62 L.Ed.2d 210 (1979), have suggested a two-tier takings analysis in a non-physical invasion case:
Where a statute deprives an individual of all his rights in specific property, the Court is likely to find a compensable taking .... When a statute takes only some of an owner’s rights, the Court will give particular weight to the magnitude of the economic harm and to the reasonableness of the owner’s investment-backed expedition of the security that property rights give.
Note, 94 Harv.L.Rev., supra at 1634 (footnotes omitted). Under this analysis, section 522(f)(2) fails to pass even the first tier. Id. Because the section completely destroys the creditor’s property rights in specific collateral, its retrospective application violates the takings clause of the Fifth Amendment.
The majority also errs in its conclusion that because the Government activity is not in the nature of a physical invasion, it is not a taking. First, I note that the section deprives the creditor of his state law rights to take possession of the collateral on default. Thus the section can be characterized as working a physical invasion. Cf. Kaiser Aetna, 444 U.S. at 179-80, 100 S.Ct. at 392-93 (taking of the right to exclude the public from private property constituted a physical invasion, and hence a Fifth Amendment violation). In this light the section is again clearly unconstitutional un*471der the per se takings analysis recently employed in Loretto v. Teleprompter Manhattan CATV Corp., - U.S. -, 102 S.Ct. 3164, 73 L.Ed.2d 868 (1982). Second, even if one accepts the majority’s nonphysical invasion characterization, the takings question does not depend on a narrow interpretation of a physical seizure of property. The hallmark is the deprivation of the former owner. United States v. General Motors Corp., 323 U.S. 373, 378, 65 S.Ct. 357, 359, 89 L.Ed. 311 (1945).
Nor is it necessary that the property inure to the benefit of the Government. See, e.g., id.; Pennsylvania Coal Co. v. Mahon, 260 U.S. 393, 413, 43 S.Ct. 158, 159, 67 L.Ed. 322 (1922) (benefit of statute inured to private homeowners); Devines v. Maier, 665 F.2d 138, 141-42 (7th Cir. 1981).
In sum, if I were required to reach the constitutional question, a course I find imprudent, I would be forced to conclude that retrospective application of section 522(f)(2) violates the takings clause of the Fifth Amendment, and is therefore invalid. On that basis, and on the basis that the court erred in even reaching the constitutional question, I respectfully dissent.
. tore Gifford, 669 F.2d 468 (7th Cir. 1982).
. The Court specifically noted that the right of the mortgagee to insist on full payment before giving up his security was the “essence” of the mortgage, 295 U.S. at 580, 55 S.Ct. at 859, and that the Act as applied diminished that right by taking the following property rights under then-existing state law:
(1) the right to retain the lien until the debt it secured is paid;
(2) the right to realize upon the security by a judicial sale;
(3) the right to determine when such a sale shall be held;
(4) the right to bid in and purchase the property at the sale;
(5) the right to control the property during the period of default.
Id. at 594-95, 55 S.Ct. at 865-66.
. The majority opinion thoroughly defends the section against a substantive due process attack, although Thorp has not pressed this argument on appeal. The decline of this now weak and disfavored doctrine is thoroughly documented, see, e.g., L. Tribe, American Constitutional Law, § 8-7 (1978); McCloskey, Economic Due Process and the Supreme Court, 1962 Sup.Ct.Rev. 34, and I see little point in belaboring this man of straw.
. Nor is the fact that the majority of bankruptcy courts dealing with the question have decided in favor of retrospective application. I note that fully half of those courts have found the statute unconstitutional if so applied — most without seeking an alternative, constitutionally acceptable construction.
Furthermore, none of those courts were aware, as the majority concedes, of the changes in preliminary drafts of the new Act reflecting Congressional intent to avoid the Radford problem by limiting retroactive application of the section. Finally, several of these courts were swayed by the argument that a “statutory gap” would be created if the new Act did not apply to post-enactment cases. Such a gap would only exist if a provision similar to § 522(t)(2) had existed under the old bankruptcy act. Prior to October 1, 1979, there was no such provision, however, and thus giving § 522(f)(2) prospective effect only, simply means that the power of avoiding liens under its provisions will take effect on October 1, 1979, and apply to all liens arising after that date. Liens created prior to that date will continue to be effective under state law — -precisely as they were under the old Bankruptcy Act and precisely as they would be if the bankruptcy proceedings had been instituted prior to enactment of the new Act but not completed until after the effective date. I therefore decline to follow the analysis, or lack thereof, of those cases, in favor of the better-reasoned approach typified by cases such as Malpeli v. Beneficial Finance Co., 7 B.R. 508 (Bkrtcy.N.D.Ill.1980).
. The majority characterizes this legislative history as “unconnected to any constitutional limitations upon Section 522(f).” The first references to Radford cited above occur in the Senate and House Committee Reports accompanying section 522(c), another subsection of the section containing the lien avoidance provisions at issue in this case. S.Rep.No. 95 -989, 95th Cong., 2d Sess. 76, reprinted in [1978] U.S.Code Cong. & Ad.News 5787, 5862; H.R. Rep.No. 95-595, 95th Cong., 2d Sess. 361, reprinted in [1978] U.S.Code Cong. & Ad.News 5963, 6317. Section 522(c) insulates exempt property from prepetition claims with certain exceptions. The primary exceptions are tax liens, alimony and support claims, and valid liens of a prepetition creditor. In referring to this final exception, the Reports adopt the rule of Long v. Bullard, 117 U.S. 617, 6 S.Ct. 917, 29 L.Ed. 1004 (1886), which held that a secured creditor’s interest in exempt property was not disturbed by bankruptcy proceedings. The Reports cite Radford in support of the Long rule. Radford relied on Long to support its holding that the bankruptcy power does not extend to the taking of private property interests. This is hardly a “technical bankruptcy rule,” as the majority suggests: rather, it is specific authority in direct support of the proposition for which it is cited above — that Congress recognized constitutional limits on its power to extinguish valid liens. It begs the question to say that because this bit of legislative history is not contained in the discussion of subsection 522(f) it is irrelevant. Delimiting our scrutiny of Congressional consideration of the limits on its bankruptcy powers to the boundaries of a particular subsection by ignoring the legislative history of a contemporaneously enacted and contiguous subsection on another aspect of the same subject (which is therefore in pari materia) does a disservice to attempts to ascertain what the Congressional intent was.
The other Congressional references to the continuing vitality of Radford occur in the Reports accompanying section 361 of the new Act, and explicitly note “the fifth amendment protection of property interests as enunciated by the Supreme Court.” S.Rep. 95-989 at 49, 5835; H.R.Rep. 95 -595 at 339, 6295 (citing Radford). Section 361 deals with methods for providing adequate protection of a creditor’s property interest in debtor property under subsequent sections which govern the grant of an automatic stay on proceedings against the property of the debtor; the use, sale, or lease of such property; and the trustee’s use of the debtor’s property to obtain credit. Again, while this section is, of course, not directly at issue here, it is in pari materia with section 522(f), and the expression of Congressional intent it evinces is relevant to the question now before us.
. As to liens perfected between the enactment date (November 6, 1978), and the effective date (October 1, 1979) of the new Act, 1 would follow the reasoning of the Court of Appeals for the Ninth Circuit in Webber v. Credithrift, 674 F.2d 796 (9th Cir. 1982) (since such “gap” creditors had notice of the provisions of the new law, their property interests were subject to the lien avoidance provisions of the new Act.).
. I simply cannot accept the majority’s conclusion that this forces “significant abandonment of the Congressional purpose.” The Congressional goal ot providing debtors a fresh start would be limited only in those cases where the security interest attached prior to November 6, 1978, the enactment date of the new Act. Most contracts of this sort are of fairly short duration — in this case, 36 months. Thus full implementation of the Congressional goal would be delayed only in a limited class of cases, and only for a fairly brief period beyond the enactment date.
The majority suggests that because Wisconsin law permits use of the same security agreement to secure additional extensions of credit, preenactment creditors could evade the effect of § 522 indefinitely. Further extensions of credit would, however, be based on a new contractual obligation, and therefore subject to the Congress’ bankruptcy power. See Kuehner, 299 U.S. at 451 52, 57 S.Ct. at 301-02. Regardless of the form of such future transactions, their effect would be to create new security interests, and as such they would be subject to § 522(f)(2), the same as any other post-enactment security interest.
. In Loretto v. Teleprompter Manhattan CATV Corp., - U.S.--, 102 S.Ct. 3164, 73 L.Ed.2d 868 (1982), the Court ruled that when physical invasion, whether by the Government or a private party, rises to the level of permanent physical occupation, a per se taking violative of the Fifth Amendment has occurred and no further recourse to the factors listed in the text is necessary.
. While the majority contends that Radford has been “greatly weakened,” it pointedly avoids the question whether the Radford takings holding remains intact. I am persuaded, see supra, that its authority remains valid. Rodrock, 642 F.2d 1193, 1197; Note, 94 Harv.L.Rev. at 1624; Note, 15 Ind.L.Rev. 593 passim.
. Thus a loan for seed or tools secured by a lien on real estate would result in the security interest attaching not to the seeds or tools or the fruits of their use, but rather to the land of the borrower, who may or may not use them on that property.
. The “home furnishings needed by the consumer for ordinary living” referred to in the majority opinion in this case include two television sets, a tape recorder, and an oak poker table. The characterization of these items as necessary, even in the American scene, would seem to place a tortured meaning on “needed.”
. The majority’s contention that the Wisconsin state law distinction between purchase money and nonpurchase money liens is evidence that the latter are not property interests is a nonsequitur. That distinctions may be made among classes of secured creditors does not obliterate the more fundamental distinction between secured and unsecured creditors. The majority’s reliance on Dames & Moore v. Regan, 453 U.S. 654, 101 S.Ct. 2972, 69 L.Ed.2d 918 (1981), is also misplaced. Although that Court found that contingent conditional attachments did not constitute compensable property interests, it also suggested that suspension of claims, rights considerably more inchoate than Thorp’s here, could give rise to a takings cause of action in the Court of Claims. Id at 688-90, 101 S.Ct. at 2291-92; see id. at 690-91, 101 S.Ct. at 2292-93 (Powell, J., concurring and dissenting) (relying on Radford for proposition that any attachment entitling a creditor to resort to specific property for the satisfaction of a claim is a property right compensable under the Fifth Amendment).
. My doubt stems from the likelihood that most bank officers, having an awareness of community relations, would be reluctant to put on the street a family which had been occupying the homestead. On the same basis, taking possession of the family heirloom jewelry or shares of stock, the sale of which would result in large capital gains to the registered holder, would be pursued only as a last resort. Here too, it seems, the threat of taking possession is often the true value of a lien or collateral.
. The majority opinion would discount that expectation by charging Thorp with actual knowledge of the terms of the new statute prior to its enactment. In light of the turbulent history of the drafting of the Act, which was the result of “eight years of study and debate,” and an extensive “process of drafting and redrafting,” In re Smith, 640 F.2d 888, 889 (7th Cir. 1981), including deletion of the only provisions affirmatively requiring retrospective application of section 522(f)(2), I find imputing such prescience unfair, even to “a sophisticated commercial creditor.”
Nor is it appropriate to say that Thorp should have been on notice at the time it entered the security transaction that its rights were contingent and subject to future changes in the law. While such reasoning might indeed be applicable to contract rights, as the cases relied upon by the majority demonstrate, secured creditors such as Thorp had every reason to believe that future bankruptcy laws not only would not, but could not, deprive them of the value of their vested property rights in specific collateral in such unprecedented fashion. The Prima case relied upon by the majority does not support its position: it involved the contractual question of reordering of priorities, not the destruction of a vested property right. Nor does the dictum from Wright v. Union Central, 304 U.S. at 516, 58 S.Ct. at 1033, provide succor to the majority position. In a later case between the same parties, the Court demonstrated that it *469did not mean that dictum literally when it established tlie rule that a secured creditor is constitutionally entitled to the value of his collateral. Wright v. Union Central Life Ins. Co., 311 U.S. 273, 61 S.Ct. 196, 85 L.Ed. 184 (1940); Note, 94 Harv.L.Rev., supra, at 1635 n.121.
. As was pointed out to Congress in hearings on the new Act, one likely result of § 522(f)(2) is to dry up or increase the cost of credit to the consumer. Bankruptcy Act Revision: Hearings on H. R. 31 & H. R. 32 Before the Sub-comm. on Civil & Constitutional Rights of the House Comm, on the Judiciary, 94th Cong., 2d Sess. 1256 (1976) (Statement of Robert Ward), reprinted in 5 A. Resnick & E. Wypyski, Bankruptcy Reform Act of 1978, A Legislative History at 1256; see Radford, 295 U.S. at 595 & n.27, 55 S.Ct. at 866 & n.27. Counsel for Thorp represented at oral argument that this is precisely what has happened. For preenactment debtors such as the Giffords who gave liens under the old bankruptcy code, but declared bankruptcy under the new Act, this has resulted in a windfall. They obtained the benefit of easy credit and better terms prevailing under the old Code, without putting their property at the same risk as other preenactment debtors, and they obtained the benefit of lien avoidance under the new Act without paying the same cost of credit as other post-enactment debtors. I simply cannot believe Congress intended such debtors to eat their cake and have it too in this fashion, and find this further evidence that Congress intended section 522(f)(2) to apply prospectively only.
. The fact, particularly in this era of economic distress and increasing unemployment, that the liening of household goods may be the only practicable method of borrowing by the Giffords and their ilk casts some doubt on the wisdom of section 522(f) in view of the fact that this same borrowing power has been virtually extinguished. This, however, is a matter of policy for the Congress to determine as it may well be doing in the light of some effects that the Act in application has wrought. Figures distinguishing purchase money from non-purchase money loans are not available.
. It is thus distinguishable from the transferable development rights referred to by the majority, which the parties in Penn Central conceded had value. Penn Central, 438 U.S. at 129, 98 S.Ct. at 2661 62.
. The majority’s benefits/burdens analysis simply proves too much. Every taking by the Government at least arguably redistributes the benefits and burdens of economic life. Pennsylvania Coal Co., 260 U.S. at 415, 43 S.Ct." at 160. The critical issue of when that regulation exceeds its constitutional bounds is properly determined by the Supreme Court-mandated analysis undertaken above, not by resort to the conclusory incantation of alliterative slogans.